QCD at High Energies at HERA

QCD at High Energies QCD at High Energies

at HERAat HERA

Max Klein

DESY ZeuthenICHEP 04 Beijing

21st of August, 2004

•The Quark Structure of the Proton

•The Strong Coupling Constant and xg

•Heavy Flavour Production

•Diffractive ep Scattering

•New Ideas and Developments

•First Results at HERA II and Outlook

s

,

W

,W±

sxyqQ 22

Deep Inelastic Scattering and Photoproduction (Q2~0)

ep-collider expts H1, ZEUS @319GeV and polarised target expt HERMES @7GeV

H1

HERMES

ZEUS

A summary of recent QCD results obtained by H1, ZEUS and HERMES

mAImAI

mAscmL

ePonpolarisati

TeVEGeVEEs

GeVEGeVE

pe

spec

ftepe

pe

100...60,50...20

1016...4

5.0...0...5.0)(:

1.543192

920,6.27

21229

PETRA

Volker Soergel and the Minister of Science of Germany, Heinz Riesenhuber, at DESY (Hamburg) announcing on 6th of April 1984 that HERA will be built.

Twenty years ago …

quarks are pointlike down to proton radius/1000

2422 10 GeVMQ ZCCNC

Neutral Current

ChargedCurrent

Xepe

Xpe

Neutral current

Charged current

1. The Quark Structure of the Proton

ZEUS r < 0.85 H1 r < 1.0 10-18m

3/)2(

)()(

)(~),(

3

222

322

vvZ

du

NC

duxxF

DDxeUUxeF

xFyfFQx

115)( pbpeL 1110)( pbpeL

e-

e+

Reduced neutral current scattering cross section at large x

Z exchange enhances electron proton cross section and reduces positron proton cross section at large Q2

sdD

bcuU

vCC xuDxyxU 2)1(~

vCC xdyxDyUx 22 )1()1(~

HERA can disentangle parton distributions

at large Q2 and large x > 0.01 within single experiments, independently of nuclear corrections and free of higher twists

e-

e+

Reduced charged current scattering cross section

- H1 and ZEUS parton distributions are in agreement - HERA experiment’s fits agree with global fits

- Gluon at low x and Q2 not well constrained

- Treatment of systematic, model and theoretical errors subject to conventions

Parton distributions unfolded with H1 data and with ZEUS data only

comparison to global fits

svv

s

xgdudu

xgAVxgDDUU

,,,

,,,,,,

QCD fits parameterise initial PDFs

H1

ZEUS

HERMES: (semi-) inclusive polarised eN

using polarised gas target internal to HERA-e

eXDep

Kph

ehXDep

,

,,

,

qeg

qqq

qeF

qqq

q

q

21

21

2

1

2

1Measure double spin asymmetriesto resolve valence and, for the first time, sea spin composition

/,),(),(

),(),(222

222

1 hhqq

hqqh Ez

QzDQxqedz

QzDQxqedzA

RICH particle ID

p/GeV10

][radc

4

0.1

0.2

Polarized quark distributions

Valence quarksdetermine p spin

Sea polarisationresolved: small,also for strange

Precision on F2 in bulk region 2-3%

2. The Strong Coupling Constant and the Gluon Distribution

][ln 22

2 FPgPQ

Fs

low x high x

ep eX: measure scattered e and Xcontrol systematics (Ee,h scales to 0.3-2%)

still potential for increased accuracy

} }

xgQ

Fs

22

ln

Derivative constrains gluon to ~20% at low x

Trijets in deep inelastic scattering new results

Interesting results also on subjets [ZEUS] and on event shapes [ZEUS+H1]

1

22

82

6.004.0

500010

pbL

y

GeVQ

Dijets in direct photoproduction novel NLO fit

Data were essential to improve determinationof strong coupling constant in “ZEUS+jets” fitusing NLO calculation (Frixione, Ridolfi). QCD fit includes also DIS jet data using DISENT (Frixione, Seymour)

003.01186.0 s

hep-ph/0407067 B.Allanach … P.Zerwas

)(005.0(exp)0011.01186.0)(.)( 2 thyMprelHERA Zs

1)((exp)0016.01160.0)(

)((exp)0015.01209.0)(0058.00046.0

2

0048.00049.0

2

HthyM

ZEUSthyM

Zs

Zs

S.Moch, J.Vermaseren, A.Vogt

Towards highest precision NNLO calculation challenges experiment

• Largest possible coverage of x, Q2 plane• Improved precision• Higher statistics at high x• Fully exploit data: jets, charm

• measure FL

Current HERA + FT DIS data on F2

hep-ph/0403192 & 0404111

H1 inclusive NC+CCZEUS inclusive NC+CC & jets

• Gluon distribution in the proton being pinned down: scaling violations, charm, jets, FL• HERA QCD fits, due to the wide range and accuracy, resolve correlation of • xg is NOT an observable. Charm treatment important (ZEUS: VFNS RT, H1: FFNS)• In the region of low x and Q2 ~ 1 GeV2 the gluon distribution becomes very small transition from hadronic to partonic behaviour at about 0.3 fm

sxg

FL data point to positive gluon distribution in the transition region

xgF

FyfF

sL

Lr

)(2

Polarised gluon distribution ΔG

Some sign that ΔG > 0 more data needed …

COMPASS, RHIC… Future requires luminous collider

polarised data also sensitive to αS

BB : Blümlein, Böttcher, GRSV: Glück,Reya,Stratman,VogelsangAAC : Asymmetry Analysis CollaborationGRV : Glück, Reya, Vogt

?

3. Heavy flavour physics at HERA hcDqbbgpq )()()(

• Fraction of c,b to inclusive F2 F2c, F2

b

• Treatment of c,b in QCD evolution : extrinsic or intrinsic, heavy or light?

• Parton radiation (DGLAP vs CCFM)

• Fragmentation functions – universal?

• Gluon in the proton

• Heavy quark and gluon content of the photon

Heavy flavours in photoproduction

Boson-Gluon fusion• evolved test of QCD at NLO [+jets, diffraction] γp: FMNR (Frixione, Mangano, Nason, Ridolfi)

DIS: HVQDIS (Harris, Smith)

Impact parameter tagging of beauty

Classic techniques: D* and

Si-vertex detectors :H1: CST: charm and beauty ICHEPZEUS: new MVD sees b’s at HERA II

)(reltp

H1

D*±

Heavy flavour identification at HERA

)(/2 S

Charm fragmentation in ep scattering

Fragmentation fractions f(cD), f(cΛc)

also determined. Agree/compete with e+e- universal behaviour of charm fragmentation

cs DDDD ,,,, *0

• Inclusive measurements consistent with NLO QCD• Discrepancies when associated with jets.• Large theoretical uncertainties (scale)• Data accurate to 10-30% can constrain PDF fits

Inclusive charm production in deep inelastic scattering

)(9

4,

),(24

22

2ccxF

Q

QxF

dxdQ

d ccccXcecep

First measurementof bottom structurefunction, uses b lifetime tagging.

Charm F2 data

with D* (ZEUS) andtagging (H1) agree. Reach now high Q2

14.57 pbL

b

c

Inclusive beauty production in deep inelastic scattering

Charm is 20% of F2.

Beauty only 2% belowvalence quark region

• Direct production dominant, similar to behaviour of charm • Resolved component measured, perhaps larger than NLO QCD• Can lead to determination of the beauty content of the photon (c.f. charm)

H1ZEUS

Beauty dijets in photoproduction

Summary of beauty data from HERA vs NLO QCD

Data of increased accuracy are above but not inconsistent with QCD

4. Hard Diffractive ep Scattering

),,,(1 2)4(

242QtxF

QdQdtddx

dIP

D

IP

NCdiff

• Why does the p sometimes remain intact? • Understand nature of diffractive exchange• Does diffraction affect p PDF’s [Martin et al]• Is diffractive exchange universal, ep – pp?• 2 g exchange high gluon density – unitarity?• Study an old phenomenon at hard scales!

Cross section factorises into coefficientfunctions and diffractive parton distributions

ZEUS: diffractive CC event

~10% of NC DIS events have gap between p and central tracks. Measure gap or detect p with LPS/VFPS

HERA allows detailed, quantitative studies. Many new results presented to ICHEP04 (inclusive, resolved y, CC, charm, jets..)

First observation by ZEUS and H1 of diffractionin charged current scattering at high Q2: 2-3%

} MX

e p

•New diffractive DIS data tagged with:

•Large rapidity gap i.e. forward detector veto.•Tagged proton using Forward p Spectrometer FPS (H1) Leading p Spectrometer LPS (ZEUS)

•Good agreement between all data

22 5.6

4.0

GeVQ

0

0.1

0.001 0.1IPx

])([ )3()3(2

)3( DL

DIP

DrIP FyfFxx

uses Regge flux (‘resolved Pomeron model’)

dzgz

dzQzzg

][

),( 2

•Extract diffractive PDFs from NLO fit to inclusive diffractive structure functions •Momentum distribution of quarks and gluons in the ‘Pomeron’: gluons dominate at large z > 0.01 unlike the non diffractive xg.

•QCD evolution (DGLAP) fits recent F2D data up

to Q2=2000 GeV2.

•If factorisation holds, these PDFs are universal and NLO QCD should describe diffractive final states and Tevatron data

z in QPM

Diffractive parton distributions

factorisation holds in DIS – jets, D*

Diffractive D*

Final states in diffractive deep inelastic scattering

Kaidalov et al.: predicted suppression of only the resolved part

resolved γ

In photoproduction need factor of ~2 suppression of NLO theory to describe the data, both in the resolved region, which is similar to pp where a factor of ~10 is needed, and in the direct region which resembles DIS

Final states in diffractive photoproduction

5. New Ideas and Developments

Low x parton radiation: forward particle production (in p direction).

How are partons (gluons) emitted?

kt ordered

angular ordered

x ordered

•DGLAP(Dokshitzer-Gribov-Lipatov-Altarelli-Parisi) DISENT/NLOJET

•BFKL(Balitsky-Fadin-Kuraev-Lipatov) ARIADNE (colour dipole. random in kt)

•CCFM(Ciafaloni-Catani-Fiorani-Marchesini) CASCADE

xjet = Ejet/Eproton >>xBj enhances BFKL effectE2

T,jet Q2 suppress DGLAP evolution

r2 =ET

2 =45 GeV

CTEQ6M

• Standard NLO pQCD prescription poor at lowest x for jets in forward direction where scale uncertainty is largest (higher orders? different radiation mechanism? best described by Ariadne – CDM – “BFKL like”)

Forward jet production in deep inelastic scattering

[interesting azimuthal (de)correlations. Also: kt dependent (“unintegrated”) pdf’s] no time

Deeply Virtual Compton Scattering – Generalised Parton Distributions

x x

2)'( qqt

2)'( kkq

access to parton correlation functions and to angular momentum of partons

~~2 ,,),,,,( EHEQtxH

q

qqqq

q

q

tEHxdxLJ

qxH

xqxH

1

1

~

0),(2

1

2

1

)0,0,(

)()0,0,(

similar to diffractive vector meson production

map transverse proton size bymeasuring t

epep

)( **22

BHDVCSBHDVCSBHDVCS AAAAAA

tbePDFGPD

Beginning of GPD phenomenology.Low x description also with colourdipole models.Need to measure t dependence: more data needed:

Hard QCD process as δ large c.f. WJ )/(

Prospects for collider experiments H1 + ZEUS - Beam charge and spin asymmetries - Tag forward protons (H1 VFPS,FPS) - Higher statistics at HERA II

126 pbL150 pbL

DVCS cross section measurementsAssume :

Prospects for fixed target experiment HERMES - Beam charge and spin asymmetries - Tag recoiling proton - Higher statistics at HERA II

)Re()cos(

)Im()sin(

DVCS

DVCS

A

A

Beam spin asymmetry

Beam charge asymmetry

DVCS asymmetry measurements

6. First Results from HERA II HERA II:

• detector and luminosity upgrade

• large, unexpected backgrounds

• identified and overcome in 2002/03

• efficient data taking since 10/03

• long run period scheduled till 2007

• polarised electron/positron p data with spin rotators at the 3 IR’s

• first data presented to ICHEP04 [HERMES: first measurement of the transverse spin structure of p hep-ex/0408013 PRL submitted]

cross section measurements with HERA II data from 2004

ZEUS H1Xepe Xpe

4.0eP

224

2

221

2

1

22

2

22,

/103.01~][1)(

)(

)()(

)(

GeVQF

FaPP

P

P

qqvexFqqexF

FPaFP

Z

er

r

qqZ

q

Ze

eNCr

parity violation ~aevq

at very high Q2

needs still larger lumi.

1

1

4.16)4.0(

1.14)32.0(

pbPL

pbPL

ZEUS

pbsysstaPeXpe

)(6.1)(8.12.0)1(

4/4.52 dof

In the SM LH coupling is excluded unless RH currents exist

Expect zero cross section at P=-1and linear dependence on P

HERA II: can now prescribepositron beam helicity also inep collider mode

Polarisation dependence firmly established for the first time.

(remember CHARM M.Jonker et al, PL 86(1979)229)

- combined H1 and ZEUS- result consistent with 0

XPFe )(

bcdudu vv ,,,,,

parton luminosities highest precision

gluonHiggs and any QCD

Low x parton radiation

DGLAP – CCFM - BFKL?QGP, astrophysics

diffraction

saturation of xgHiggs

sunification

HERA is an important part of HEP and is doing well again

HERA & LHC

Outlook

Gluon and quark distributions essential to measure the Higgs cross section

•HERA experiments submitted ~100 papers to Beijing.

•With thanks to all the HERA activists.

•Thanks to the organisers and speakers.

•Special thanks to

Elke-Caroline Aschenauer,

Mandy Cooper-Sarkar, Karin Daum, Didar Dobur, Claudia Glasman, Claire Gwenlin, Delia Hasch, Ewelina Lobodzinska, Uta Stößlein

and to a few more men for help with this presentation.